INDEX

Explanations

references to a bar or related activities

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "bar" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lihood

-0.85

UAL

-0.71

wip

-0.63

ATK

-0.61

 Incarn

-0.60

IBLE

-0.60

 autom

-0.60

 Scarlet

-0.58

 ozone

-0.58

 Vehicles

-0.58

POSITIVE LOGITS

itone

1.42

bers

1.35

iatric

1.35

celona

1.29

becue

1.22

bell

1.21

bara

1.20

ber

1.18

onial

1.15

bec

1.13

Activations Density 0.033%

references to a bar or related activities

occurrences of the word "bar" in various contexts

No Comments

No Known Activations